Neddylation inhibition ameliorates steatosis in NAFLD by boosting hepatic fatty acid oxidation via the DEPTOR-mTOR axis

Neddylation inhibition ameliorates steatosis in NAFLD by boosting hepatic fatty acid oxidation via the DEPTOR-mTOR axis Marina Serrano-Maciá 1 , Jorge Simón 1 , 2 , Maria J. González-Rellan 3 , 4 , Mikel Azkargorta 5 , Naroa Goikoetxea-Usandizaga 1 , Fernando Lopitz-Otsoa 6 , Diego Saenz De Urturi 7 , Rubén Rodríguez-Agudo 1 , Sofia Lachiondo-Ortega 1 , Maria Mercado-Gomez 1 , Virginia Gutiérrez de Juan 6 , Maider Bizkarguenaga 6 , David Fernández-Ramos 2 , 6 , Xabier Buque 7 , 8 , Guido A. Baselli 9 , 10 , Luca V.C. Valenti 9 , 10 , Paula Iruzubieta 11 , 12 , Javier Crespo 11 , 12 , Erica Villa 13 , Jesus M. Banales 7 , 14 , Matias A. Avila 7 , 15 , Jose J.G. Marin 7 , 16 , Patricia Aspichueta 6 , 8 , James Sutherland 17 , Rosa Barrio 17 , Ugo Mayor 18 , 19 , Félix Elortza 5 , Dimitris P. Xirodimas 20 , Rubén Nogueiras 3 , 4 , Teresa C. Delgado 1 , ** , María Luz Martínez-Chantar 1 , 2 , * ABSTRACT Objective: Neddylation is a druggable and reversible ubiquitin-like post-translational modification upregulated in many diseases, including liver fibrosis, hepatocellular carcinoma, and more recently, non-alcoholic fatty liver disease (NAFLD). Herein, we propose to address the effects of neddylation inhibition and the underlying mechanisms in pre-clinical models of NAFLD. Methods: Hepatic neddylation measured by immunohistochemical analysis and NEDD8 serum levels measured by ELISA assay were evaluated in NAFLD clinical and pre-clinical samples. The effects of neddylation inhibition by using a pharmacological small inhibitor, MLN4924, or molecular approaches were assessed in isolated mouse hepatocytes and pre-clinical mouse models of diet-induced NAFLD, male adult C57BL/6 mice, and the AlfpCre transgenic mice infected with AAV-DIO-shNedd8. Results: Neddylation inhibition reduced lipid accumulation in oleic acid-stimulated mouse primary hepatocytes and ameliorated liver steatosis, preventing lipid peroxidation and inflammation in the mouse models of diet-induced NAFLD. Under these conditions, increased Deptor levels and the concomitant repression of mTOR signaling were associated with augmented fatty acid oxidation and reduced lipid content. Moreover, Deptor silencing in isolated mouse hepatocytes abolished the anti-steatotic effects mediated by neddylation inhibition. Finally, serum NEDD8 levels correlated with hepatic neddylation during the disease progression in the clinical and pre-clinical models 1 Liver Disease Lab, Center for Cooperative Research in Biosciences (CIC bioGUNE), Basque Research and Technology Alliance (BRTA), 48160, Derio, Bizkaia, Spain 2 Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBERehd), Carlos III National Health Institute, Madrid, Spain 3 Department of Physiology, Center for Research in Molecular Medicine and Chronic Diseases (CIMUS), University of Santiago de Compostela-Instituto de Investigación Sanitaria, Galician Agency of Innovation (GAIN), Xunta de Galicia, 15782, Santiago de Compostela, Spain 4 Centro de Investigación Biomédica en Red de Fisiopatologia de la Obesidad y Nutrición (CIBERobn), Carlos III National Health Institute, Madrid, Spain 5 Proteomics Platform, Center for Cooperative Research in Biosciences (CIC bioGUNE), Basque Research and Technology Alliance (BRTA), Carlos III Networked Proteomics Platform (ProteoRed-ISCIII), 48160, Derio, Bizkaia, Spain 6 Precision Medicine and Metabolism Lab, Center for Cooperative Research in Biosciences (CIC bioGUNE), Basque Research and Technology Alliance (BRTA), 48160, Derio, Bizkaia, Spain 7 Department of Physiology, Faculty of Medicine and Nursing, University of the Basque Country UPV/EHU, 48940, Leioa, Bizkaia, Spain 8 Biocruces Bizkaia Health Research Institute, 48903, Barakaldo, Bizkaia, Spain 9 Department of Pathophysiology and Transplantation, Università degli Studi di Milano, 20122, Milan, Italy 10 Department of Transfusion Medicine and Hematology eTranslational Medicine, Fondazione IRCCS Ca’Granda Ospedale Maggiore Policlinico, 20122, Milan, Italy 11 Gastroenterology and Hepatology Department, Marqués de Valdecilla University Hospital, 39008, Santander, Spain 12 Clinical and Translational Digestive Research Group, Research Institute Marqués de Valdecilla (IDIVAL), 39011, Santander, Spain 13 Department of Gastroenterology, Azienda Ospedaliero- Universitaria & University of Modena and Reggion Emilia, 41121, Modena, Italy 14 Department of Liver and Gastrointestinal Diseases, Biodonostia Health Research Institute, Donostia University Hospital, University of the Basque Country (UPV/EHU), Ikerbasque, 20014, San Sebastian, Spain 15 Hepatology Programme, Center for Applied Medical Research (CIMA), IDISNA, University of Navarra, 31008, Pamplona, Spain 16 Experimental Hepatology and Drug Targeting (HEVEFARM), IBSAL, University of Salamanca, 37007, Salamanca, Spain 17 Ubiquitin-likes And Development Lab, CIC bioGUNE, Basque Research and Technology Alliance (BRTA), 48160, Derio, Bizkaia, Spain 18 Department of Biochemistry and Molecular Biology, Faculty of Science and Technology, University of the Basque Country (UPV/EHU), 48940, Leioa, Spain 19 Ikerbasque - Basque Foundation for Science, Bilbao, Spain 20 Centre de Recherche en Biologie cellulaire de Montpellier (CRBM-CNRS), Univ. Montpellier, UMR5237, Montpellier 34090, Cedex 5, France *Corresponding author. CIC bioGUNE, Ed. 801A Parque Tecnológico de Bizkaia, 48160, Derio, Bizkaia, Spain. Fax: þ34 944 061301. E-mail: [email protected] (M.L. Martínez-Chantar). **Corresponding author. CIC bioGUNE, Ed. 801A Parque Tecnológico de Bizkaia, 48160, Derio, Bizkaia, Spain. Fax: þ34 94406130. E-mail: [email protected] (T.C. Delgado). Received February 24, 2021  Revision received June 9, 2021  Accepted June 14, 2021  Available online 18 June 2021 https://doi.org/10.1016/j.molmet.2021.101275 Original Article MOLECULAR METABOLISM 53 (2021) 101275 Ó2021 The Authors. Published by Elsevier GmbH. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). www.molecularmetabolism.com 1 Conclusions: Overall, the upregulation of Deptor, driven by neddylation inhibition, is proposed as a novel effective target and therapeutic approach to tackle NAFLD. Ó2021 The Authors. Published by Elsevier GmbH. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Keywords Neddylation; NAFLD; Deptor; mTOR; Fatty acid oxidation; MLN4924 1. INTRODUCTION Non-alcoholic fatty liver disease (NAFLD), or (as recently redefined) metabolic dysfunction associated fatty liver disease (MAFLD) [1], comprises a group of conditions that all share the excessive accumulation of fat in the liver as a common feature. NAFLD ranges from simple steatosis and non-alcoholic steatohepatitis (NASH) to advanced fibrosis, eventually leading to cirrhosis and Hepatocellular Carcinoma (HCC). Patients with NAFLD have a higher risk of liver-related morbidity and mortality, along with metabolic comorbidities [2e5]. Today, NAFLD is the most common liver disorder in the Western world, as its global prevalence is estimated to be 24% [6]. Despite the alarming numbers associated with the pathology of NAFLD worldwide and the large number of emerging therapies currently undergoing clinical trials, there remains no approved pharmacotherapy for the clinical management of NAFLD. To date, ongoing studies on NAFLD therapies have been primarily designed to target four main pathways: 1) hepatic fat accumulation; 2) oxidative stress, inflammation, and apoptosis; 3) intestinal microbiomes and metabolic endotoxemia; and 4) hepatic fibrosis. Though most clinical trials have focused on monotherapies, the use of a single drug acting on multiple pathways may offer an attractive therapeutic strategy in combatting NAFLD. Post-translational modifications (PTMs) affect many biological processes. Therefore, a single pharmacological inhibitor can alter the activity, localization, or function of several proteins and simultaneously target multiple deregulated pathways. Post-translational modifications refer to the covalent and commonly enzymatic modification of proteins after biosynthesis. PTMs are an essential mechanism needed to diversify protein functions and dynamically coordinate their signaling networks. On this basis, the study of protein expression and PTMs, along with their role in NAFLD development and progression, has attracted attention in the postgenomic era. For example, the high-throughput profiling of PTMs has the potential to provide a molecular classification of fatty liver; changes to liver protein acetylation, phosphorylation, and OeN-ace- tylglucosamine glycosylation were found to be altered in NAFLD states [7]. More recently, the focus has been geared toward the mediation of PTMs by ubiquitin and ubiquitin-like proteins in NAFLD. Furthermore, several authors have shown how ubiquitin-specific peptidases and some ubiquitin ligases (E3) can regulate the development of NAFLD [8e12]. NEDD8 (neural precursor cell-expressed, developmentally down-regulated protein8) is an evolutionarily conserved 8-kD protein closely related to ubiquitin, and like ubiquitin, it can be conjugated to the specific lysine residues of target proteins in eukaryotes ea process called neddylation. Though the conjugation of NEDD8 to its targets is usually associated with the stabilization of the protein, it has also been implicated in the activity of its target proteins. Despite the conjugation of NEDD8 onto the cullin subunit of CRLs (Cullin-RING ubiquitin ligase) promoting its assembly and activating its ubiquitin ligase activity being the best-characterized NEDD8 target [13], the number of newly identified neddylated targets has been increasing in the last several years with the development of novel techniques [14,15]. Therefore, the impact of neddylation on diseases can be much more extensive. It has been recently determined that neddylation is increased in advanced liver fibrosis and HCC [16,17]. Other authors have shown that the degradation of the splicing factor SRSF3 (serine rich splicing factor 3) in a neddylation-dependent mechanism is disrupted in early metabolic liver disease and may contribute to NASH progression, cirrhosis, and ultimately HCC [18]. Finally, after treatment with a pharmacological inhibitor of neddylation, obese mice present decreased weight gain and, consequently, an improved liver phenotype [19]. Herein, we have further addressed the relevance of hepatic neddylation in NAFLD and the therapeutic efficacy of neddylation inhibition in cell models in vitro and mouse models of diet-induced NAFLD. We provide evidence that hepatic neddylation inhibition decreases liver steatosis by boosting fatty acid oxidation in a process partly mediated by the impaired mammalian target of rapamycin (mTOR) signaling, as regulated by DEPTOR (DEP-domain containing mTOR-interacting protein). Moreover, we have identified that the levels of NEDD8 in serum appear to correlate with NAFLD disease progression and reversal upon dietary interventions or neddylation inhibition. Overall, treating NAFLD by Abbreviations NAFLD non-alcoholic fatty liver MAFLD metabolic dysfunction associated fatty liver disease NASH non-alcoholic steatohepatitis PTMs Post-translational modifications NEDD8 neural precursor cell-expressed, developmentally down-regulated protein8 HCC Hepatocellular carcinoma mTOR mammalian target of rapamycin DEPTOR DEP-domain containing mTOR-interacting protein NRF2 nuclear factor erythroid 2erelated factor 2 CRLs Cullin-RING ubiquitin ligase FAO fatty acid oxidation DNL de novo lipogenesis VLDL very-low-density lipoprotein OXPHOS oxidative phosphorylation ETC electron transport chain ROS reactive oxygen species MCDD methionine choline-deficient diet CDHFD choline-deficient high-fat diet FA fatty acid OA oleic acid OCR oxygen consumption rate Tnf tumor necrosis factor Ccl-2 chemokine (CeC motif) ligand 2 Il-6 interleukin-6 Il-1 b interleukin-1 beta Timp-2 tissue inhibitor of metalloproteinases 2 Original Article 2MOLECULAR METABOLISM 53 (2021) 101275 Ó2021 The Authors. Published by Elsevier GmbH. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). www.molecularmetabolism.com targeting neddylation may be a fast and effective strategy to regulate altered signaling pathways and metabolic reactions. 2. METHODS 2.1. Human samples The human samples used in this work were obtained after the informed consent of the patients and in accordance with the Ethical Code of the World Medical Association and Helsinki Declaration and upon approval of the Hospital Marques de Valdecilla ethics Committee. Four well-characterized cohorts of NAFLD patients, obtained from the University-Hospital Marques de Valdecilla (Santander, Spain), were used. Liver biopsies were assessed and classified by an expert pathologist. In order to study global neddylation levels in liver tissue by immunohistochemistry, two cohorts of NAFLD patients, including lean (n ¼4) and obese patients (n ¼15), were used. Conversely, another two cohorts of patients were analyzed to evaluate NEDD8 levels in serum (n ¼19-lean and n ¼24-obese; Suppl. Table 1). The inclusion criteria established to develop the current project were as follows: age between 18 and 75 years old, no other acute or chronic disease except for type 2 diabetes and obesity (metabolic syndrome) in accordance with the physical and biochemical analysis, and alcohol consumption should be less than 30 g/day for men and 20 g/day for women. The exclusion criteria established were viral infection or liver disease caused by drugs (drug-induced liver injury). All recruited patients were of Caucasian origin. The control groups consisted of healthy human liver and serum samples (n ¼7 liver samples for histological study, and n ¼8 serum samples from women and men between 40 and 70 years old). Additional protocols used are provided in the supplemental material. The mRNA levels of all genes involved in the neddylation pathway were analyzed from liver biopsies obtained during bariatric surgery from a cohort of 125 NAFLD patients, described earlier [20]. Individuals with increased alcohol intake (>30/20 g/day in M/F), viral and autoimmune hepatitis, or other causes of liver disease were excluded. Informed consent was obtained from each patient, and the study protocol was approved by the Ethical Committee of the Fondazione IRCCS Ca’ Granda and conformed to the ethical guidelines. Briefly, RNA was isolated from liver biopsies (miRNeasy mini-kit, Qiagen) and sequenced using a HiSeq 4000 (Illumina). Reads were aligned and counted (GRCh37, ENSEMBL v75 reference), employing STAR and RSEM packages. Count normalization was performed using DESeq2. The differential expression of the genes of interest was assessed by a two-tailed Wilcoxon Rank Sum test. Statistical analyses were conducted using R 3.5.3 software, and p <0.05 was considered statistically significant. 2.2. Animal maintenance, diets, and treatment The animal procedures were performed in accordance with the European Research Council for animal care and use and the National Institute of Health guide for care and use of Laboratory animals. The maximal authority of the Country Council of Bizkaia and the Institutional Animal Care and Use Committee of CIC bioGUNE approved the animal procedures. Adult (three-month-old) male C57BL/6J mice were acquired from Charles River Laboratories and accommodated into the AALAC-accredited CIC bioGUNE animal facilities and maintained at 211  C, 45 10% humidity, and a 12/12 h light/dark cycle. Animal maintenance was based on ad libitum access to water and the respective diet. The control group was fed a standard chow diet with 0.3% methionine and 1.030 mg/kg of choline (Teklad Global 14% Protein Rodent Maintenance diet; Envigo 2014C). Dietary-mouse models of NAFLD included: a high-fat diet (HFD; D12492, Research Diets), choline-deficient, high-fat diet (CDHFD; D05010402, Research Diets), and 0.1% methionine and choline-deficient diet (0.1% MCDD; A02082006i, Research Diets). A group of 5 animals, randomly divided, was treated with MLN4924 (MeDChemExpress, MCE) dissolved in a vehicle solution of 2-hydroxypropyl- b -cyclodextrin (SigmaeAldrich) by oral gavage or subcutaneous injection at a dose of 60 mg/kg during the last two weeks of treatment in case of the animals fed 0.1% MCDD and during the last three weeks in the CDHFD model. A control group of 5 animals was treated with a vehicle solution. After 4 weeks of 0.1% MCDD and 6 weeks of CDHFD, animals were sacrificed. In addition, a group of 10 Alfp-Cre mice was used to silence Nedd8 specifically in the liver and were randomly divided into control or silenced groups. These animals expressed Cre-recombinase open reading (ORF) under the control of albumin and a -fetoprotein, abundant in the hepatocytes [21]. They were infected with adenovirus in the tail vein, which contains a short hairpin Nedd8 RNA sequence and double floxed sequence that allows for the recombination with Cre and, therefore, the insertion of the sequence (AVV-DIO-shNedd8). The infection was performed 3 weeks before the start of the 6 weeks of CDHFD. Additional protocols used are provided in the supplemental material. Finally, adult (three-month-old) male C57BL/6J mice were maintained during 3 weeks on a 0.1% methionine, choline-deficient, high-fat diet (CDAAHFD; A06071302, Research Diets) and after this period, switched back to a standard chow diet for one additional week. For all experiments, the liver and blood were extracted and maintained at 80  C or formalin-fixed solution until further analysis. 2.3. Primary mouse hepatocytes and treatments Male adult mice (C57BL/6J) were acquired from Charles River Laboratories and accommodated into the CIC bioGUNE animal facility. After liver perfusion and 3 h attachment, primary mouse hepatocyte media were changed to 0% FBS MEM (Gibco) þPSG (Gibco) overnight. The following morning, hepatocytes were incubated for 6 h with either 400 m M oleic acid (SigmaeAldrich) conjugated with BSA-fatty acidfree or BSA-fatty acid-free diluted in 0% MEM. Neddylation inhibition was achieved using 3 m M of MLN4924 in DMSO, an inhibitor of the Nedd8 Activating Enzyme E1 Subunit 1 (MeDChemExpress). In some experiments, 20 m M of Etomoxir (SigmaeAldrich) was used during the 6 h period to inhibit fatty acid oxidation. Primary mouse hepatocytes were transfected by 12 h incubation with 100 nM Deptor siRNA (Thermo Fisher, Silencer Select siRNA DEPTOR, sequence Sense: 5 0 - AGACGGCGAUAAAACUCAUTT- 3 0 and Anti-sense: 5 0 - AUGAGUUUUAUCGCCGUCUCT e3) to evaluate the role of DEPTOR in fatty acid oxidation. Finally, Nedd8 was silenced using siNEDD8 to evaluate the physiological role of neddylation inhibition (Sigma siNEDD8 Sequence sense: 5 0 - CAUCUACAGUGGCAAGCAA [dT][dT] -3 0 and 5 0 - UUGCUUGCCACUGUAGAUGAG [dT][dT]- 3 0 ). The reagent used to silence DEPTOR and Nedd8 in hepatocytes was Lipofectamine 2000 (Invitrogen, Ref: 1168e019), according to the user’s manual. SilencerÔNegative Control No. 1 siRNA (Thermo Fisher) was used as a negative siRNA transfection control. Additional protocols used are provided in supplemental material. 2.4. Statistical analysis Prism 8 (GraphPad Software, version 8.4.0) was used to perform statistical analyses. A one-way analysis of variance (ANOVA) followed by Tukey (comparing all pairs of columns) was used for three or more groups, while Student’s t-test was used for 2 groups. Grubbs’test was performed to determine the significant outliers. A p <0.05 was MOLECULAR METABOLISM 53 (2021) 101275 Ó2021 The Authors. Published by Elsevier GmbH. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). www.molecularmetabolism.com 3 considered statistically significant. Statistical parameters are reported in the figure legends. 3. RESULTS 3.1. Hepatic neddylation is augmented in clinical and pre-clinical NAFLD Previous evidence from our laboratory has shown that NEDD8- mediated modifications are increased in patients with advanced liver fibrosis and cirrhosis and mouse models of liver fibrosis [16]. Herein, we aimed to extend our studies to assess the role of neddylation in the early stages of NAFLD. On this basis, hepatic global neddylation levels, evaluated by immunohistochemistry (IHC), were increased in liver biopsies from a cohort of well-characterized NAFLD patients, lean and obese, in comparison to age-matched healthy controls (Figure 1A, Suppl. Table 1). In-depth analyses revealed that hepatic global neddylation levels positively correlate with the NAS score (Figure 1B). Then, we evaluated the levels of hepatic neddylation in mouse models of diet-induced NAFLD, reflecting different stages of the pathology: i) animals on a high-fat diet (HFD) for 20 weeks, characterized by steatosis and ballooning; ii) animals on a choline-deficient, high-fat diet (CDHFD) for 6 weeks and characterized by mild steatosis and inflammation; and iii) animals on a 0.1% methionine and cholinedeficient diet (0.1% MCDD) for 4 weeks and characterized by enhanced macro-steatosis and overt inflammation. Hepatic global neddylation was induced in all animal models of NAFLD (Figure 1CeE), with higher increases corresponding to more aggressive dietary interventions. Further longitudinal hepatic characterization of steatosis, ballooning, inflammation, and fibrosis levels in the dietary mouse models of NAFLD is shown in Suppl. Fig. 1aec. The deregulation of the neddylation pathway does not occur at the transcriptional level in the clinical setting or our animal models of diet-induced NAFLD, as no significant changes in the mRNA levels for E1, E2, and E3 or the deconjugating enzymes for the NEDD8 pathway were observed (Suppl. Fig. 2aec). In summary and concurrent with early evidence [19], hepatic neddylation was augmented in patients suffering from NAFLD and mouse models of diet-induced NAFLD. 3.2. Neddylation inhibition reduces lipid accumulation in NAFLD pre-clinical models The effects of the specific small pharmacological inhibitor of neddylation targeting the NEDD8 activating enzyme (NAE1), MLN4924, ([(1S,2S, 4R)-4-[4-[[(1S)-2,3-dihydro-1H-inden-1-yl]amino]pyrrolo [2,3-d]pyrimidin-7-yl]-2 hydroxycyclopentyl]methyl sulfamate) (commercialized as Pevonedistat) [22], were evaluated in hepatocyte cell models of steatosis. Treatment with MLN4924 and reduced neddylation, assessed as decreased neddylated cullins in isolated mouse hepatocytes stimulated with oleic acid (OA), significantly reduced lipid content, as determined by bodipy staining (Suppl. Fig. 3a,Figure 2A). Though MLN4924 has been previously shown to induce tumor cell death, in primary mouse hepatocytes, Figure 1: Neddylation is augmented in clinical and pre-clinical mouse models of non-alcoholic fatty liver disease (NAFLD).A. Hepatic neddylation levels assessed by immunohistochemistry and B. Pearson correlation and coefficient between hepatic neddylation and NAS score in NAFLD patients (n ¼19) in comparison with a group of healthy controls (n ¼7). Hepatic neddylation levels assessed by immunohistochemistry in healthy animals (n ¼5) and C. animals maintained during 20 weeks on a high-fat diet (HFD) (n ¼11), D. animals maintained during 6 weeks on a choline-deficient high-fat diet (CDHFD) (n ¼8) and E. animals maintained during 4 weeks on a 0.1% methionine- and choline-deficient diet (0.1% MCDD) (n ¼6). Scale bar corresponds to 100 m m. Data are shown as average SEM and Student’s t-test was used to compare groups. *p<0.05 and ***p<0.001 are shown. Original Article 4MOLECULAR METABOLISM 53 (2021) 101275 Ó2021 The Authors. Published by Elsevier GmbH. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). www.molecularmetabolism.com Figure 2: Neddylation inhibition reduces lipid accumulation in non-alcoholic fatty liver disease (NAFLD) pre-clinical models.A. and B. Bodipy staining in primary mouse hepatocytes isolated from wild-type mice and treated with MLN4924 3 m M(n¼4) or siRNA against Nedd8 (n ¼3) in the presence of oleic acid (OA) during 6h. Sudan Red staining and quantification of total hepatic triglycerides in C. animals maintained during 6 weeks on choline-deficient and high-fat diet (CDHFD) (n ¼5) and animals on CDHFD and treated with MLN4824 (CDHFD þMLN4924) (n ¼5), in D. animals maintained during 4 weeks on 0.1% methionine- and choline-deficient diet (0.1% MCDD) (n ¼4) and treated with MLN4924 (0.1%MCDD þMLN4924) (n ¼4) and finally, in E. Alfp-Cre/AAV-DIO-shNEDD8 mice were maintained for 6 weeks on CDHFD (n ¼5). Transaminase serum levels of ALT (alanine transaminase) and AST (aspartate transaminase) in healthy animals and F. animals maintained during 6 weeks on CDHFD and CDHFD þMLN4924, G. animals maintained during 4 weeks on 0.1% MCDD and 0.1% MCDD þMLN4924, and in H. the hepatic silenced NEDD8 mice model (Alfp-Cre/AAV-DIO-shNEDD8) maintained during 6 weeks on CDHFD. Scale bar corresponds to 100 m m. Data are shown as average SEM and One-way ANOVA test was used to compare groups. *p<0.05, **p<0.01 and ***p<0.001 are shown. MOLECULAR METABOLISM 53 (2021) 101275 Ó2021 The Authors. Published by Elsevier GmbH. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). www.molecularmetabolism.com 5 MLN4924 does not induce significant caspase-3 activity, an apoptosis marker (Suppl. Fig. 3b), suggesting that the observed effects in decreased lipid content do not reflect cell death. The lack of MLN4924 associated death-induced effects in steatotic hepatocytes also indicates reduced adverse effects related to MLN4924 treatment. Similar findings were obtained after Nedd8 silencing in isolated mouse hepatocytes using molecular approaches, such as siRNA (Figure 2B; Suppl., Fig. 3c,d), suggesting that the anti-steatotic effects ascribed to MLN4924 are related to neddylation inhibition and not unspecific off-target effects. Moreover, the effect of neddylation inhibition was evaluated in 2 mouse models of NAFLD, shown earlier to have high Nedd8 levels (Figure 1D,E). In the CDHFD-fed mouse, animals were fed ad libitum for 3 weeks to induce mild steatosis, and afterward, MLN4924 treatment was given (60 mg/kg), once a week for other 3 weeks, totaling in 6 weeks of dietary intervention. In the 0.1% MCDD model, MLN4924 treatment (60 mg/kg) was given every 4 days, beginning after the second week of the diet, a time point characterized by steatosis and inflammation, and the experiment finished at 4 weeks of diet (Suppl. Fig. 4a). MLN4924 treatment and associated hepatic neddylation inhibition after CDHFD and 0.1% MCDD significantly decreased hepatic steatosis, quantified by Sudan red staining and biochemically measuring hepatic triglycerides (Figure 2C and D, and Suppl. Fig. 4b,c). We silenced Nedd8, specifically in the hepatocytes of CDHFD animals, by using Alfp-Cre/AAV-DIO-shNedd8 to confirm further that there are no off-target effects associated with MLN4924 treatment (Suppl. Fig. 4a). Under these conditions, Nedd8 silencing, specifically in the hepatocytes in vivo, is associated with reduced hepatic neddylation levels and liver steatosis (Figure 2E, Suppl. Fig. 4d). Supporting these data, serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were significantly reduced after neddylation inhibition, indicating diminished liver injury in the 0.1% MCDD model. In contrast, no changes were found in ALT or AST in the CDHFD model compared with control animals (Figure 2FH). Of note, the final body weight in MLN4924-treated mice was not altered, relative to the vehicle-treated animals on NAFLD diets (CDHFD: 32.5 0.9 g and CDHFD þMLN4924: 31.2 1.2 g, p-value ¼0.57; 0.1% MCDD: 23.7 0.7 g and 0.1% MCDD þMLN4924: 24.3 1.1 g, p-value ¼0.63; CDHFD: 33.1 0.6 g and CDHFD þshNedd8 32.1 1.3g, p-value ¼0.45), suggesting that the anti-steatotic effects of neddylation inhibition are not related to decreased adiposity. In summary, neddylation-specific inhibition shows anti-steatotic effects in isolated mouse hepatocytes and pre-clinical mouse models of NAFLD. 3.3. mTOR inhibition via DEPTOR accumulation plays a role in neddylation inhibition mediated anti-steatotic effects in NAFLD preclinical models Recent reports have highlighted that the ATF/CREB family is involved in regulating many aspects of lipid metabolism processes, including regulating critical enzymes and regulators involved in lipogenesis, fatty acid oxidation, and lipoprotein metabolism [23]. mRNA levels of members of the ATF/CREB family, such as Atf4 and 6(Activating Transcription Factor 4 and 6) and Crebh (CAMP Responsive Element Binding Protein), show no significant regulation as a result of the MLN4924 treatment in the CD-HFD and 0.1% MCDD NAFLD mouse models (Suppl. Fig. 5a,b). Thus, we performed high-throughput proteomics Liquid Chromatography-Mass Spectrometry (LC-MS)-based analyses in animals fed CDHFD for 6 weeks and 0.1% MCDD for 4 weeks to better understand NAFLD at the proteome level. Heat map representation shows the top 50 altered proteins after CDHFD and 0.1% MCDD (Figure 3A; Supplemental File 1). A file containing all identified proteins is provided in Supplemental File 1. Ingenuity pathway analysis (IPA) was used to identify the major canonical pathways involved in NAFLD, suggesting that eukaryotic initiation factor 2 (EiF2) signaling, which integrates a diverse array of stressrelated signals to regulate both global and specific mRNA translation, and lipid metabolism, are profoundly altered. Moreover, the mTOR pathway, which plays an essential role in regulating lipid metabolism [15,24,25] , is also highly altered in NAFLD (Figure 3B). In addition, proteomics LC-MS-based analysis was also performed to assess the effect of MLN4924 treatment in the dietary mouse models of NAFLD (Figure 3C; Supplemental File 1). IPA analysis aimed at identifying potential upstream regulators that explain the changes observed in the proteomics panel as a result of MLN4924 treatment was performed. The common upstream regulators identified in the two dietary models studied include MYC, HNF4A, TP53, and POR (Figure 3D). Further studies are necessary to understand the upstream regulators involved in neddylation triggering inhibition. mTOR signaling plays a vital role in regulating lipid metabolism, such as lipogenesis and oxidative fluxes in the liver [26]. In addition, other studies have shown that in tumor cells, DEPTOR is degraded by the SCF b TrCP (Skp1-Cullin-F box proteins) E3 ubiquitin ligase, the founding member of CRLs, resulting in mTOR activation and cell proliferation [27]. Importantly, SCF b TrCP E3 ubiquitin ligase is activated through NEDD8 conjugation of the conserved lysines of cullins. Herein, we show that the neddylation inhibition in vivo in CDHFD and 0.1% MCDD induced protein DEPTOR, and as a consequence of the concomitant mTOR inhibition, the phosphorylation of S6 protein (pS6), a downstream target of mTOR, was reduced without changes in Deptor gene expression (Figure 4A and B). Likewise, neddylation inhibition using MLN4924 pharmacological treatment in OA-stimulated hepatocytes increased DEPTOR content (Figure 4C). Under these conditions, when silencing Deptor by using siRNA-based molecular approaches in primary mouse hepatocytes, MLN4924 treatment did not significantly reduce the cellular lipid content (Figure 4CeE). Thus, mTOR inhibition via DEPTOR accumulation plays a role in the neddylation inhibition-mediated anti-steatotic effects. 3.4. Neddylation inhibition boosts fatty acid oxidation coupled with oxidative phosphorylation in NAFLD pre-clinical models The regulation of fat homeostasis in the liver results from a balance between uptake, anabolism, catabolism, and export from the liver. Indeed, a tight regulation between fat uptake, de novo lipogenesis (DNL), fatty acid degradation/oxidation (FAO), and triglyceridecontaining very-low density lipoproteins (VLDL) export occurs in the liver. Both unaltered changes in weight gain and serum triglycerides and cholesterol after MLN4924 treatment or Nedd8 silencing in vivo in CDHFD- and 0.1% MCDD-fed animals (Suppl. Fig. 6aec) strongly suggest that neither fat uptake nor VLDL export, respectively, are altered after neddylation inhibition. Furthermore, we know from previous evidence that DNL is reduced due to consuming methionine- and choline-deprived diets and is, therefore, not a central player in the mouse models of dietary-induced NAFLD [28,29]. mTOR signaling plays an essential role in regulating lipid metabolism, such as lipogenesis and oxidative fluxes in the liver. Indeed, pharmacologic intervention via mTORC1 inhibition increases the oxidation of endogenous fatty acids in the primary cultures of rat hepatocytes [30] and ameliorates hepatic steatosis in chronic-binge ethanol-fed mice [31]. MLN4924 treatment to isolated mouse hepatocytes stimulated by OA induces FAO activity (Figure 5A). Increased FAO under these Original Article 6MOLECULAR METABOLISM 53 (2021) 101275 Ó2021 The Authors. Published by Elsevier GmbH. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). www.molecularmetabolism.com circumstances most likely accounts for the augmented fueling of the TCA cycle by acetyl-CoA facilitating the transference of reducing equivalents from various intermediates to coenzymes and the production of NADH and FADH. These two reduced coenzymes fuel the respiratory chain. In agreement, neddylation inhibition by MLN4924 significantly induced oxidative phosphorylation (OXPHOS), an electron transport-linked phosphorylation, and ATP-linked respiration in mouse hepatocytes, measured by Seahorse-based analysis (Figure 5B). Furthermore, in the presence of etomoxir, a well-described inhibitor of FAO, MLN4924 was unable to decrease the hepatic lipid content of mouse hepatocytes stimulated by OA (Figure 5C), further confirming that the pharmacological inhibition of neddylation, through MLN4924, may be exerting its lipid-lowering effect by increasing FAO. Likewise, FAO activity was shown to be induced in MLN4924-treated CDHFD and Figure 3: Liver proteomics characterization by Liquid Chromatography-Mass Spectrometry (LC-MS) in mouse models of non-alcoholic fatty liver disease (NAFLD) and treated with MLN4924.A. Heatmaps showing the top 50 significantly different most differentially expressed proteins and B. Ingenuity Pathway Analysis (IPA) of top canonical pathways in animals maintained during 6 weeks on choline-deficient and high-fat diet (CDHFD) as well as in animals maintained during 4 weeks on 0.1% methionine- and cholinedeficient diet (0.1% MCDD). C. Heatmaps showing the top 50 significantly different most differentially expressed proteins; D. Ingenuity Pathway Analysis (IPA) of Upstream Regulators in animals maintained during 6 weeks on choline-deficient and high-fat diet (CDHFD) and animals on CDHFD and treated with MLN4824 (CDHFD þMLN4924), as well as in animals maintained during 4 weeks on 0.1% methionine- and choline-deficient diet (0.1% MCDD) and animals on 0.1% MCDD and treated with MLN4824 (0.1% MCDD þMLN4924). MOLECULAR METABOLISM 53 (2021) 101275 Ó2021 The Authors. Published by Elsevier GmbH. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). www.molecularmetabolism.com 7 0.1% MCDD-fed rodents (Figure 5D and E), whereas OXPHOS, measured with a Seahorse extracellular analyzer in isolated fresh liver mitochondria, was increased after MLN4924 treatment in CDHFD-fed rodents (Figure 5F). Overall, neddylation inhibition boosted fatty acid oxidation, coupled with oxidative phosphorylation and decreasing liver fat content. 3.5. Neddylation inhibition reduces oxidative stress, lipid peroxidation, and inflammation in NAFLD pre-clinical models Fat accumulation in hepatocytes during NAFLD provides a potential substrate for lipid peroxidation, a process in which oxidants (e.g., free radicals) attack lipids containing carbonecarbon double bond(s), especially polyunsaturated fatty acids (PUFAs). Malondialdehyde and, in particular, 4-hydroxy-2-nonenal (4-HNE) are the primary omega-6 fatty acid lipid peroxidation products. Herein, mouse models of dietinduced steatosis, CDHFD, and 0.1% MCDD, show an increased accumulation of dihydroethidium (DHE), a fluorescent marker of reactive oxygen species (ROS), and 4-HNE. Importantly, both DHE and 4-HNE were significantly decreased after MLN4924 treatment in vivo in the CDHFD and 0.1% MCDD-fed rodents (Figure 6A,B). Moreover, in the CDHFD and 0.1% MCDD-fed rodents, MLN4924 treatment was associated with an induced expression of NRF2 (nuclear factor Figure 4: DEPTOR accumulation by neddylation inhibition mediates the anti-steatotic effects in pre-clinical models of non-alcoholic fatty liver disease (NAFLD). Western blot analysis of neddylated cullins, DEPTOR, phosphorylated S6 (pS6) and total S6 and beta-actin both in A. animals maintained during 6 weeks on choline-deficient and high-fat diet (CDHFD) (n ¼5), and animals on CDHFD and treated with MLN4824 (CDHFD þMLN4924) (n ¼5) as well as in B. animals maintained during 4 weeks on 0.1% methionine- and choline-deficient diet (0.1% MCDD) (n ¼5) and animals on 0.1%MCDD and treated with MLN4824 (0.1% MCDD þMLN4924) (n ¼5). C. Western blot of neddylated cullins, and DEPTOR (n ¼3), D. mRNA analysis of DEPTOR levels (n ¼3) and E. Bodipy staining of lipids in hepatocytes stimulated with OA and MLN4924 when silencing DEPTOR (n ¼3). Data are shown as average SEM. One-way ANOVA test and Student’s t-test were used to compare groups. *p<0.05, **p<0.01 and ***p<0.001 are shown. Original Article 8MOLECULAR METABOLISM 53 (2021) 101275 Ó2021 The Authors. Published by Elsevier GmbH. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). www.molecularmetabolism.com erythroid 2erelated factor 2; Suppl. Fig. 7a). NRF2 is a master regulator of antioxidative responses known to drive the expression of numerous cytoprotective genes involved in xenobiotic metabolism, antioxidant responses, and anti-inflammatory responses [32]. Thus, the expression of genes involved in the glutathione (GSH) metabolism was further assessed. In the CD-HFD and 0.1% MCDD MLN4924- treated groups, a tendential overexpression at the transcriptional level was observed for the subunits GclC and GclM of Gcl (glutamate cysteine ligase), an enzyme that catalyzes the rate-limiting step in GSH synthesis. Moreover, Gsr (glutathione reductase), involved in GSH reduction, and GstA (glutathione S-transferase), which catalyzes the conjugation of reduced GSH to a vast number of exogenous and endogenous hydrophobic electrophiles, was augmented after MLN4924 treatment (Suppl. Fig. 7b,c). The oxidative stress response induced by lipid accumulation in hepatocytes may further trigger the inflammatory cytokine responses. Indeed, though the hepatic mRNA levels of proinflammatory cytokines, such as Tnf (tumor necrosis factor), Ccl-2 (chemokine (CeC motif) ligand 2, and Il-6 (interleukin-6), were increased after CDHFD, treatment with MLN4924 reduced their levels (Figure 6C). Likewise, Tnf- a , Ccl-2, and Il-1 b (interleukin-1 beta) were increased after 0.1% MCDD for 4 weeks and decreased after neddylation inhibition in vivo with MLN4924. Moreover, Timp-2 (tissue inhibitor of metalloproteinases 2), an early pro-fibrogenic gene, is increased with 0.1% MCDD and Figure 5: Neddylation inhibition boosts fatty acid oxidation coupled with oxidative phosphorylation in non-alcoholic fatty liver disease (NAFLD) pre-clinical models.A. Fatty acid oxidation (FAO) fluxes (n ¼3); and B. Oxygen consumption rate (OCR) as analyzed using a Seahorse analyzer in primary mouse hepatocytes stimulated with oleic acid (OA) during 6 h in the presence and absence of MLN4924. C. Immunofluorescence staining of Bodipy with or without stimulation with etomoxir, an inhibitor of FAO in primary mouse hepatocytes isolated from wild type mice and treated with MLN4924 in the presence of OA during 6h (n ¼3e4). D. and E. Fatty acid oxidation fluxes in animal models maintained during 6 weeks on choline-deficient and high-fat diet (CDHFD), and animals on CDHFD and treated with MLN4824 (CDHFD þMLN4924) as well as in animals maintained during 4 weeks on 0.1% methionine- and choline-deficient diet (0.1% MCDD) and animals on 0.1%MCDD and treated with MLN4824 (0.1% MCDD þMLN4924) (n ¼5 per group). F. OCR as analyzed using a Seahorse analyzer in isolated liver mitochondria from mice maintained during 6 weeks on CDHFD or CDHFD þMLN4924. Scale bar corresponds to 100 m m. One-way ANOVA test and Student’s t-test, respectively, were used to compare groups. # p ¼0.09, *p<0.05, **p<0.01 and ***p<0.001 are shown. MOLECULAR METABOLISM 53 (2021) 101275 Ó2021 The Authors. Published by Elsevier GmbH. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). www.molecularmetabolism.com 9